CN112392658A - Method for detecting irregular turbine operation using direct and indirect wind speed measurements - Google Patents

Abstract

The invention relates to a method for detecting irregular turbine operation using direct and indirect wind speed measurements. Method for operating a wind turbine, the wind turbine comprising a wind characteristic sensor for measuring a wind characteristic and at least one wind turbine status sensor for measuring a status of the wind turbine, the method comprising: determining or adjusting 102 one or more wind characteristic relationships; and performing 104 an operational phase comprising: measuring wind characteristics by using a wind characteristic sensor, thereby obtaining measured wind characteristics; measuring a state of the wind turbine with at least one wind turbine state sensor and determining an estimated wind characteristic from the measured state of the wind turbine and a parameter of the wind turbine; comparing the estimated wind characteristic to an expected wind characteristic determined from the measured wind characteristic, wherein the expected wind characteristic is determined based on one or more wind characteristic relationships; and operating or shutting down the wind turbine based at least in part on the comparison.

Description

Method for detecting irregular turbine operation using direct and indirect wind speed measurements

Technical Field

The subject matter described herein relates to methods for operating a wind turbine, and to a wind turbine, and more particularly to methods for operating a wind turbine comprising a wind characteristic sensor for measuring a wind characteristic and at least one wind turbine state sensor for measuring a state of the wind turbine, an estimate of the wind characteristic being obtained from the state of the wind turbine.

Background

A wind turbine typically includes a tower and a nacelle mounted on the tower. The rotor is rotatably mounted to the nacelle and is coupled to a generator by a shaft. A plurality of blades extend from the rotor. The blades are oriented such that wind passing over the blades turns the rotor and rotates the shaft, thereby driving the generator to produce electricity.

Wind turbines convert wind energy into mechanical energy, for example into rotational kinetic energy, and the mechanical energy is further converted into electrical energy, typically by a wind turbine generator. In order to control the forces and/or torques acting on the blades, the blade pitch angle, i.e. the angle of attack of the blades of the rotor of the wind turbine with respect to the direction of the wind flow, may be adjusted. Thus, the rotational speed of the rotor of the wind turbine and the electrical power generated by the wind turbine generator driven by the rotor through the shaft of the wind turbine may be controlled by adjusting the pitch angle of the blades of the wind turbine.

For one or more blades of a wind turbine, the blade pitch angle may be adjusted for each blade individually or collectively. As wind speed changes, the blade pitch angle of one or more blades of the wind turbine is adjusted to maintain the rotor speed and torque within operational limits for maximizing the efficiency of the wind turbine generator to produce electrical energy while minimizing the risk of damaging the wind turbine due to, for example, sudden wind gusts.

The wind turbine may reach a stall condition, i.e. a condition such that the maximum power produced by the wind turbine starts to decrease if the angle of attack of one or more blades increases. For practical wind conditions, a further increase in angle of attack results in a decrease in power for the one or more blades at which a stall condition is created. The minimum angle of attack that produces a stall condition is referred to as the critical angle of attack for the actual wind conditions.

The wind turbine may be operated in a stall condition, but when the angle of attack of one or more blades is further increased, a significant stall condition or a deep stall condition may result. Operating a wind turbine in a significant stall condition or a deep stall condition is undesirable.

Typical critical angles of attack are in the range of 15 to 20 degrees. In general, if the angle of attack exceeds a critical angle, the wind turbine is said to be in a stall condition, i.e. is stalling. To avoid any stall on parts of the blade, the angle of attack typically needs to be 3 to around 5 degrees lower than the critical angle of attack during operation of the wind turbine. Thus, a significant stall condition or a deep stall condition may be any condition in which the wind turbine is stalling when the angle of attack exceeds a critical angle of attack.

Under significant stall conditions, the turbulence of the wind flow may cause chaotic or irregular dynamics of the wind flowing at the wind turbine. Operation under significant stall conditions may be part of wind turbine operation, but significant stall conditions are generally undesirable due to the prevalent negative effects, such as, for example, chaotic or irregular wind flow and/or power reductions. Moreover, excessively high wind speeds or wind gusts may damage the wind turbine, and operation in the presence of strong winds at significant stall conditions may pose a significant risk of damage to the blades and/or other wind turbine components.

A fault or disturbance condition of a wind turbine may be caused by different causes, such as, for example, icing of blades of the wind turbine, dust deposited on blades of the wind turbine, aging of wind turbine components, or other external or internal factors affecting the function of the wind turbine.

Accordingly, it would be beneficial to reliably detect and/or prevent a significant stall condition of a wind turbine or a fault or disturbance condition of a wind turbine.

Disclosure of Invention

According to one aspect, there is provided a method for operating a wind turbine, the wind turbine comprising a wind characteristic sensor for measuring a wind characteristic and at least one wind turbine condition sensor for measuring a condition of the wind turbine, the method comprising: determining or adjusting one or more wind characteristic relationships; and an execution phase, the operation phase comprising: measuring wind characteristics by using a wind characteristic sensor, thereby obtaining measured wind characteristics; measuring a state of the wind turbine with at least one wind turbine state sensor and determining an estimated wind characteristic from the measured state of the wind turbine and a parameter of the wind turbine; comparing the estimated wind characteristic to an expected wind characteristic determined from the measured wind characteristic, wherein the expected wind characteristic is determined based on one or more wind characteristic relationships; and operating or shutting down the wind turbine based at least in part on the comparison.

Accordingly, the present disclosure is directed to accurately measuring wind characteristics of wind present at a wind turbine, such as wind speed and/or direction and/or shear, the presence of turbulence in the wind flow, and the like. To do so, the wind characteristics are measured with a wind characteristics sensor. Further, the measurement of the state of the wind turbine is performed with at least one wind turbine state sensor, which state may for example comprise the speed of the rotor of the wind turbine and/or the torque of the rotor and/or the generated power of the wind turbine.

According to a further aspect, there is provided a wind turbine comprising: at least one wind measurement sensor; and a wind turbine state sensor to measure a state of the wind turbine for estimating wind characteristics at the wind turbine location; a control system configured to control the wind turbine based at least in part on an input formed by a measured wind characteristic measured by the wind measurement sensor and a measured wind turbine state measured by the wind turbine state sensor.

A method for operating a wind turbine comprising a wind characteristic sensor for measuring a wind characteristic and at least one wind turbine condition sensor for measuring a condition of the wind turbine, the method comprising:

determining or adjusting one or more wind characteristic relationships; and

performing an operational phase, the operational phase comprising:

measuring the wind characteristic with the wind characteristic sensor, thereby obtaining a measured wind characteristic;

measuring the state of the wind turbine with the at least one wind turbine state sensor and determining an estimated wind characteristic from the measured state of the wind turbine and a parameter of the wind turbine;

comparing the estimated wind characteristic to an expected wind characteristic determined from the measured wind characteristic, wherein the expected wind characteristic is determined based on the one or more wind characteristic relationships; and

operating or shutting down the wind turbine based at least in part on the comparison.

Solution 2. the method according to solution 1, wherein determining or adjusting one or more wind characteristic relationships is performed when the wind turbine is not in a significant stall condition and not in a disturbance condition, and comprises:

measuring the wind characteristic of the wind turbine with the wind characteristic sensor of the wind turbine, thereby obtaining a measured wind characteristic of the wind turbine;

measuring the state of the wind turbine with the at least one wind turbine state sensor and determining an estimated wind characteristic of the wind turbine from the measured state of the wind turbine and a parameter of the wind turbine,

determining or adjusting a relationship between the measured wind characteristic of the wind turbine and the estimated wind characteristic of the wind turbine; and

adjusting the one or more wind characteristic relationships to include the relationship between the measured wind characteristic of the wind turbine and the estimated wind characteristic of the wind turbine.

Technical solution 3. the method according to technical solution 1 or 2, wherein determining or adjusting one or more wind characteristic relationships comprises:

operating a same type of wind turbine as the wind turbine when the same type of wind turbine is not in a significant stall condition and is not in a disturbance condition, the same type of wind turbine including a wind characteristic sensor and at least one wind turbine condition sensor; and, during the operation of the wind turbines of the same type, the method further comprises:

measuring a wind characteristic of the same type of wind turbine with the wind characteristic sensor of the same type of wind turbine, thereby obtaining a measured wind characteristic of the same type of wind turbine; and

measuring the state of the same type of wind turbine with the at least one wind turbine state sensor of the same type of wind turbine and determining an estimated wind characteristic of the same type of wind turbine from the measured state of the same type of wind turbine and a parameter of the same type of wind turbine;

determining or adjusting a relationship between the measured wind characteristics of the wind turbines of the same type and the estimated wind characteristics of the wind turbines of the same type; and

adjusting the one or more wind characteristic relationships to include the relationship between the measured wind characteristics of the same type of wind turbine and the estimated wind characteristics of the same type of wind turbine.

Solution 4. the method of any of the preceding solutions, wherein determining or adjusting one or more wind characteristic relationships comprises:

simulating wind and wind turbine operation for the wind turbine without significant stall and disturbance conditions of the wind turbine, the simulation based at least in part on a model of the wind turbine;

obtaining simulated wind characteristics, simulated states and simulated parameters of the wind turbine, determining simulated estimated wind characteristics from the simulated states of the wind turbine and the simulated parameters of the wind turbine;

determining or adjusting a relationship between the simulated wind characteristic and the simulated estimated wind characteristic; and

adjusting the one or more wind characteristic relationships to include the relationship between the simulated wind characteristic and the simulated estimated wind characteristic.

Solution 5. the method of any of claims 1 to 4, wherein the one or more wind characteristic relationships are further combined into a single combined relationship, and wherein the desired wind characteristic is based on the single combined relationship.

Solution 6. the method according to any of the preceding solutions, characterized in that the wind turbine is operated or shut down according to the desired wind characteristic determined from the measured wind characteristic, when the comparison shows that the estimated wind characteristic differs significantly from the desired wind characteristic determined from the measured wind characteristic.

Solution 7. the method of any of the preceding solutions, wherein the comparing comprises obtaining a difference between the estimated wind characteristic and the desired wind characteristic, and operating the wind turbine is based at least in part on a magnitude of the difference.

The method of claim 8, the method of claim 7, wherein the wind turbine operates based on the estimated wind characteristic when the magnitude of the difference is below a first threshold.

Claim 9. the method of claim 7 or 8, wherein the wind turbine operates based on the desired wind characteristic when the magnitude of the difference is above the first threshold.

Claim 10. the method of any of claims 7 to 9, wherein the turbine switches to a safe mode of operation or shuts down when the magnitude of the difference is above a second threshold.

Claim 11 the method of any of claims 7 to 10, wherein a message is transmitted to an operator when the magnitude of the difference is above the first threshold and/or the second threshold.

Solution 12. method according to any of solutions 7 to 11, characterized in that the magnitude of the difference is memorized at different points in time forming a sequence, and wherein a normal condition or a significant stall or disturbance condition is determined based on the sequence, and wherein in case of a significant stall or disturbance condition a type of fault is determined from the sequence, and the wind turbine is operated according to the determined type of fault.

Solution 13. the method of any of claims 1 to 12, wherein operating or shutting down the wind turbine comprises adjusting a pitch angle to avoid a significant stall condition of the wind turbine.

Claim 14 the method of any one of claims 1 to 13, wherein the wind characteristic is a wind speed, and the wind characteristic sensor measures a magnitude of the wind speed.

The invention according to claim 15 provides a wind turbine comprising:

at least one wind measurement sensor; and

a wind turbine state sensor to measure a state of the wind turbine for estimating wind characteristics at the wind turbine location;

a control system configured to control the wind turbine based at least in part on an input formed by a measured wind characteristic measured by the wind measurement sensor and a measured wind turbine state measured by the wind turbine state sensor,

wherein the control system is configured to operate the wind turbine according to the method of any of claims 1 to 14.

Further aspects, details and advantages are apparent from the following description, the accompanying drawings and the dependent claims.

Drawings

The disclosure will be explained in accordance with the following exemplary figures.

FIG. 1 illustrates a wind turbine having a nacelle, a rotor, and rotor blades according to an embodiment of the present disclosure.

FIG. 1A illustrates details of a wind turbine, particularly illustrating a wind turbine generator and a shaft of the wind turbine, according to embodiments of the present disclosure.

FIG. 1B illustrates a method for operating a wind turbine according to an embodiment of the present disclosure.

FIG. 2 illustrates determining or adjusting one or more wind characteristic relationships according to a method of the present disclosure.

FIG. 3 illustrates operational stages of a method for operating a wind turbine according to an embodiment of the present disclosure.

FIG. 4 illustrates details related to a method for operating a wind turbine according to an embodiment of the present disclosure.

Detailed Description

Reference will now be made in detail to the various embodiments, one or more examples of which are illustrated in the drawings.

FIG. 1 illustrates a wind turbine 10 including a tower 12 placed on a support system 14, a nacelle 16 with a rotor 18, the nacelle 16 being connected to a rotatable hub 20. One or more rotor blades 22 are configured to convert kinetic energy of the wind into rotational kinetic energy of rotor 18. Each blade has a blade root portion 24, a load transfer region 26, where rotation is transmitted to the rotatable hub 20 at the load transfer region 26. As the wind component flows in direction 28, the rotor and the rotatable hub rotate about an axis of rotation 30. Along the rotor blade 22, a pitch axis 34 is shown in FIG. 1.

The control system 36, which may be located at the wind turbine (as in FIG. 1) or elsewhere, is configured to control a pitch angle of the rotor blades in relation to an angle of attack with respect to a wind direction, in order to control, for example, a speed or torque of the rotor blades of the wind turbine, wherein the speed or torque is applied to the rotor by the wind. The wind turbine further has a yaw axis 38 for orienting the rotor blades about the tower 12 with respect to different wind directions. Processor 40 may be part of control system 36.

As shown in FIG. 1A, the nacelle 16 of the wind turbine further includes a wind turbine generator 42 for generating electrical energy from rotational kinetic energy of the rotor generated from kinetic energy of wind as a function of the pitch angle of the rotor blades.

In the present disclosure, it is intended that the wind characteristics may include one or more wind speeds, one or more wind shears, one or more time or spatial derivatives of wind speed, one or more wind directions. For example, the wind characteristic may be a scalar quantity related to the magnitude of the wind speed at the wind turbine location (e.g., the wind speed in the direction 28 shown in FIG. 1). The wind characteristic may also be, for example, a vector related to the wind speed at the wind turbine location, or a set of scalar or set of vectors related to one or more wind speeds at or near the wind turbine location, where the wind speed may be the wind speed at a given location in space or an average spatial or temporal wind speed at or near the wind turbine location. For example, wind characteristics may be described in terms of an ordered tuple of real numbers related to wind speed at or near the wind turbine location.

In some embodiments, the wind characteristic may be a magnitude of a wind speed, in particular a scalar wind speed or a vector describing a wind speed. For example, wind characteristics may be measured in m/s.

In FIG. 1A, additional details of wind turbine 10, and in particular details of nacelle 16 of wind turbine 10, are illustrated. In particular, rotor shaft 44 transmits kinetic energy to the wind turbine generator for generating electrical energy from the kinetic energy of the wind. The rotor shaft exhibits a longitudinal axis 45, which forms the axis of rotation of the rotor shaft. Gearbox 46 may be used to control the rotational speed and torque of a high speed shaft 48 that drives the wind turbine generator. Wind turbine generator 42 is driven by rotational kinetic energy of a high speed shaft 48 driven by rotor shaft 44 through a gearbox 46 for generating electrical energy. Thus, rotor shaft 44 transmits rotational movement to high speed shaft 48 through gearbox 46, and the rotational speed of rotor shaft 44 is typically lower than the rotational speed of high speed shaft 48. Rotor shaft 44 is coupled to blades of rotor 18 of the wind turbine and, when wind imparts rotational movement to the rotor, the rotor shaft rotates accordingly.

FIG. 1A further illustrates a coupling 50 between high speed shaft 48 and wind turbine generator 42, supports 52 and 54, a yaw drive mechanism 56, yaw drive mechanism 56 for rotating the nacelle about yaw axis 38 for orienting the rotor with respect to wind speed direction 28. Wind characteristic sensor 58 may measure a wind characteristic at the wind turbine location, such as a wind speed flowing in direction 28. The wind characteristic sensor 58 of the wind turbine may be, for example, an anemometer. In general, and without limitation to any of the other features described with respect to FIG. 1A, the anemometer of the wind turbine may be located on top of the nacelle.

As shown in FIG. 1A, the bearings 60, 62 may support a shaft or other component of a wind turbine. The wind turbine may further include a pitch assembly 66, and pitch assembly 66 may include a pitch drive assembly 68 for controlling the pitch angle of one or more blades. The assembly may include sensors 70 for one or more rotor blades, a pitch bearing 72, a pitch drive motor 74, a pitch drive gearbox 76, and a pitch drive pinion 78.

Overspeed control system 80 may be present. Further indicated in fig. 1A is a cable 82 for transmitting signals from or to the control system of the wind turbine. Finally, the actuator 84 may provide the actual pitch angle of the wind turbine blade, the blade connected to the cavity 86 presenting an inner surface 88 and an outer surface 90.

As used herein, the term "blade" is intended to be representative of any device that provides a reactive force when in motion relative to a surrounding fluid (e.g., air that creates wind at a wind turbine location). As used herein, the term "wind turbine" is intended to be representative of any device that generates rotational energy from wind energy, and more specifically, converts kinetic energy of wind into mechanical energy. A "wind turbine generator" further converts mechanical energy into electrical energy, typically by means of a wind turbine generator.

Although each commercial wind turbine is typically equipped with an anemometer on the nacelle, these anemometers are generally not used as input for turbine control because their readings are too unreliable. In contrast, some modern wind turbines use model-based estimation techniques to calculate wind speed based on the performance of the turbine itself. However, these estimators rely on accurate model information to be stored in the controller, or on assumptions about wind turbine operation or conditions affecting the wind turbine. Thus, the estimator is not usable to detect abnormal turbine operation such as icing or stall, because in these cases/situations the model parameters are no longer correct and the estimator no longer reports the correct wind speed. Furthermore, if such abnormal operation has not been detected, the controller will erroneously control the turbine, e.g. drive the turbine to a deep stall, which may cause additional losses in power production, or even damage the wind turbine or some of its components.

FIG. 1B illustrates a method 100 for operating a wind turbine according to an embodiment of the present disclosure. The method 100 for operating a wind turbine includes determining or adjusting 102 one or more wind characteristic relationships and executing 104 an operational phase.

As used herein, a sensor for measuring wind characteristics may be, in particular, a wind turbine anemometer. The method of the present disclosure allows in particular to calibrate the wind turbine anemometer or the sensor for measuring wind characteristics, and the measured wind speed of the wind turbine anemometer or the sensor for measuring wind characteristics becomes a more reliable and usable quantity for the control and monitoring of the wind turbine system. The model wind is reliable if the wind turbine is operating under normal, non-disturbance conditions. If the wind turbine is indeed operating offline or in a stall or disturbance condition, the wind speed obtained based on the model may be erroneous and thus the turbine may not be operating at its optimal operating parameters or may even be damaged.

The present disclosure provides a highly accurate redundant wind speed measurement that will be used to detect, for example, blade icing, blade failure, and other turbine anomalies or disturbance conditions that may be detected by wind speed deviations. In some embodiments, a significant stall or deep stall condition is also detectable. Wind speed measurements made by wind characteristic sensors (e.g., anemometers) are typically much more accurate during non-operational times of the wind turbine and/or during stall or disturbance conditions of the wind turbine. Thus, wind speed measured by a wind characteristic sensor or anemometer may be used during stall or disturbance conditions instead of wind speed obtained/estimated by using a model for precise control or in order to prevent damage, as long as for example systematic errors affecting the wind speed measurements obtained by the wind characteristic sensor (e.g. anemometer) are properly handled. Furthermore, this allows for possible power calculations, improved accuracy and performance of the recovery operation after e.g. calm or storm conditions etc. The method of the present disclosure also enables power curve measurements based on, for example, nacelle wind speed measurements.

For example, there are serious drawbacks when calibrating an anemometer with respect to a meteorological mast. For example, the correlation between wind characteristics at two locations is not good or not always good due to the distance between the meteorological mast and the turbine. Furthermore, this type of calibration may only be applicable to a specific wind turbine. For turbines without meteorological masts, such a calibration, e.g. obtained from another turbine, may not be applicable and/or significant variability as a function of factors such as e.g. local terrain configuration may affect the quality or reliability of the calibration. When using a wind speed measuring device (such as e.g. a lidar) that measures wind characteristics in front of the rotor, calibration may be less necessary. However, devices such as lidar tend to be expensive.

As used herein, it is intended that the state of the wind turbine may include, for example, rotor speed and/or electrical power generated by the wind turbine generator and/or torque of the rotor and/or rotor shaft. It is intended that the parameters for the operation of the wind turbine may for example comprise the pitch angle of the blades of the wind turbine or for example the torque of the generator of the wind turbine and/or the configuration of the gearbox of the wind turbine. The parameters are assumed to be known quantities.

The wind characteristics may include one or more wind speeds, one or more wind directions, one or more wind accelerations, and/or wind turbulence at one or more locations at or near the wind turbine location. It is intended that both wind characteristics and wind turbine states and parameters may be one or more scalars and/or one or more vectors describing one or more quantities.

Different types of sensors may be used to measure wind characteristics at the wind turbine location. It may be possible to use a wind measuring bar positioned at a distance from the wind turbine, e.g. upstream of the wind turbine. However, the value of the wind characteristic measured at the measuring rod may differ from the value at the wind turbine location, e.g., surrounding terrain and/or objects may produce a significant difference in the value of the wind characteristic measured at the rod relative to the wind characteristic at the wind turbine location.

When measuring wind characteristics at a wind turbine location using a local wind characteristics sensor, such as an anemometer placed at the wind turbine, the measurements are typically affected by errors (e.g., systematic errors) due to the presence of the wind turbine and the wind turbine blades themselves. Thus, due to the influence of the presence of the wind turbine and the wind turbine blades themselves, the values of the wind characteristics measured by local sensors at the wind turbine location (e.g. by an anemometer located at the wind turbine) cannot be directly used to determine the true wind characteristics at the wind turbine location.

In the nominal case, it is beneficial to use the wind turbine itself as a measurement instrument for determining the wind characteristics at the wind turbine location. Given the actual relevant operating parameters of a wind turbine (such as the pitch angle of the blades), the state of the wind turbine (e.g., rotor speed and/or power output) is nominally related to the wind characteristics at the wind turbine location (e.g., to the local wind speed). Thus, given the actual known values of the parameters of the wind turbine, it is possible to estimate the wind characteristics from the state of the wind turbine. Thus, in the nominal case, the wind turbine itself may replace the sensors used to measure the wind characteristics at the wind turbine location. However, if a stall condition (e.g. a significant stall or deep stall condition) or a disturbance condition (e.g. in the presence of ice or dirt on the blades) occurs, the wind turbine may no longer be used for estimating the wind characteristics, because for the actual values of the parameter such as the pitch angle, the correlation between the true actual wind characteristics and the state of the wind turbine becomes irregular or chaotic or unreliable or subject to significant errors.

Thus, it is beneficial to detect stall or disturbance conditions without relying on wind characteristics estimated from the state of the wind turbine, but also to avoid the situation where only a wind characteristic sensor (such as e.g. an anemometer) is used directly, as the wind characteristic sensor is typically subject to significant systematic or statistical errors.

Detecting stall or disturbance conditions is beneficial for operating the wind turbine, for example for avoiding damage to the wind turbine and/or for improving delivery of output power.

Some types of sensors (e.g., lidar) may be capable of measuring wind characteristics in the vicinity of the wind turbine that may be used to reliably determine the wind characteristics at the wind turbine location with sufficient accuracy and precision, however lidar may be expensive or at least impractical in some cases.

Accordingly, it is beneficial to calibrate a wind characteristics sensor (e.g., a local anemometer) at a wind turbine location in order to overcome systematic errors introduced by the presence of the wind turbine that typically affect the wind characteristics sensor.

Furthermore, it is advantageous to compare the wind characteristics estimated from the measured state of the wind turbine with the wind characteristics obtained by the calibrated wind characteristics sensor, taking into account the operating parameters. Under normal circumstances, the wind characteristics estimated from the measured state of the wind turbine are more reliable and accurate, but the values of the wind characteristics obtained with the calibrated wind characteristics sensor are close to the values of the estimated wind characteristics. That is, the two values are comparable, whereas without calibration the wind characteristic sensor is subject to significant errors, but under normal circumstances the estimated wind characteristic value will typically be more accurate and precise.

Under significant stall or disturbance conditions, the wind characteristics estimated from the measured state of the wind turbine may be erroneous and may differ significantly from the values of the wind characteristics obtained with the calibrated wind characteristics sensors. Thus, the values obtained by the calibrated wind characteristic sensor may be used for plausibility checking of the wind characteristic estimated from the measurement state of the wind turbine.

It is advantageous to compare the wind characteristics obtained from the measured state of the wind turbine with the wind characteristics obtained with a calibrated wind characteristics sensor, in particular in order to detect stall and disturbance conditions of the wind turbine. In particular, the comparison may be based on a difference between an estimated wind characteristic based on the measured state of the wind turbine and a wind characteristic obtained with a calibrated sensor (e.g. a calibrated anemometer) for measuring the wind characteristic. Without calibration, significant errors may affect wind characteristic sensors, such as anemometers, and thus the measurements may be erroneous.

In fig. 2, a calibration phase of a method for operating a wind turbine according to some embodiments of the present disclosure is illustrated. In particular, FIG. 2 illustrates how the relationship of one or more wind characteristic relationships may be determined or adjusted. As used herein, the term "calibration phase" may thus refer to the determination or adjustment of one or more wind characteristic relationships, i.e. the determination or adjustment 200 of one or more wind characteristic relationships is performed in the calibration phase. The wind characteristic relationship may be implemented by any data structure capable of associating information about wind characteristics with other information about wind characteristics. For example, the wind characteristic relationship may be a transfer function that associates another vector (e.g., a vector describing the estimated desired wind characteristic) with a vector (e.g., a vector describing the measured wind characteristic). The wind characteristic relationship may also be implemented as a set of ordered pairs of vectors, where for each ordered pair the first component is a vector relating to, for example, a measured wind characteristic and the second component is a vector relating to, for example, a desired estimated wind characteristic. The expected estimated wind characteristic may form an expected value of the wind characteristic estimated by using a physical model of the wind turbine, for example based at least in part on a measured state of the wind turbine.

The calibration phase for determining or adjusting one or more wind characteristic relationships may comprise: measuring 202 a wind characteristic with a wind characteristic sensor of the wind turbine (e.g., with wind characteristic sensor 58 of FIG. 1A), thereby obtaining wind characteristic data; a state of the wind turbine is measured 204 with at least one wind turbine state sensor and an estimated wind characteristic of the wind turbine is determined from the measured state of the wind turbine and a parameter of the wind turbine. The wind turbine condition sensor may, for example, specifically measure a rotational speed of a rotor shaft 44 of the wind turbine.

Further parameters, such as for example the pitch angle of one or more rotor blades, may be taken into account for determining the estimated wind characteristics. It is intended that the estimation of the wind characteristics is based on, inter alia, a physical model of the wind turbine. As shown in FIG. 2, the calibration phase may further include determining or adjusting 206 a relationship between measured wind characteristics of the wind turbine and estimated wind characteristics of the wind turbine. The relationship may be based, inter alia, on measuring wind characteristics and/or estimating wind characteristics and historical sequences of such characteristics stored in a convenient data structure, such as in a list of ordered pairs stored in a memory of, for example, control system 36 and/or processor 40. It is intended that the relationship in block 206 may be identified using, for example, a transfer function.

The wind characteristics measured by the wind characteristics sensor 58 of the wind turbine (e.g. by a local anemometer) are taken advantage of by the symbol w_MeasuringAnd (4) indicating. Respectively by means of symbols s_{Turbine wheel}Indicating the state of the wind turbine and using the symbol p_{Turbine wheel}An operating parameter of the wind turbine is indicated.

It is intended that w_Measuring、s_{Turbine wheel}、p_{Turbine wheel}May be scalar or vector. In some alternative embodiments, these quantities may alternatively refer to the same type of wind turbine or quantities related to the simulation of a wind turbine.

State s of wind turbine_{Turbine wheel}May include, for example, a rotor speed, a rotor torque, and/or, for example, a rotational speed of rotor shaft 44 of the wind turbine and/or a torque of rotor shaft 44 and/or a power output of wind turbine generator 42. It is intended that an estimation of the wind speed characteristics at the location of the wind turbine is possible when measuring the state of the wind turbine, possibly under consideration of the operational parameters of the wind turbine assuming knowledge.

Operating parameter p_{Turbine wheel}May include, for example, a pitch angle of the rotor blades, a torque parameter of the wind turbine generator, an actual configuration of the gearbox, etc.

Knowing the state s of a wind turbine_{Turbine wheel}And an operating parameter p of the wind turbine_{Turbine wheel}It is possible to estimate the wind characteristics at the location of the wind turbine. The estimated wind characteristics as a function of the state and parameters of the wind turbine are indicated by the following equation:

w_estimating＝w_Estimating(s_{Turbine wheel},p_{Turbine wheel})

It is intended that w_MeasuringAnd w_EstimatingMay be scalar or vector and they may be compared to each other using, for example, a suitable metric such as the euclidean distance between the scalars or vectors. w is a_Estimating＝w_Estimating(s_{Turbine wheel},p_{Turbine wheel}) Can be calculatedIn particular on model-based estimation techniques, and in particular on the use of physical models, for example of wind turbines and/or wind turbine components.

At s_{Turbine wheel}、p_{Turbine wheel}In some alternative embodiments involving wind turbines of the same type, w_EstimatingThe same type of wind turbine is also concerned. At s_{Turbine wheel}、p_{Turbine wheel}In some alternative embodiments involving simulation of a wind turbine, w_EstimatingIt also relates to the simulation of wind turbines.

Assuming an operating parameter p of a wind turbine_{Turbine wheel}It is known, for the sake of brevity, to state that the estimation of the wind characteristics at the wind turbine location is from the state s of the wind turbine_{Turbine wheel}Is obtained and, equivalently, can be written as w_Estimating＝w_Estimating(s_{Turbine wheel}) Implicitly assuming p pairs_{Turbine wheel}In which p is_{Turbine wheel}Are known.

When there is no significant stall or disturbance condition, w_EstimatingMay be a good estimate of the actual wind characteristics at the wind turbine location, whereas in significant stall or disturbance conditions, w_EstimatingThe true value of the wind characteristic at the wind turbine location may be significantly deviated.

On the other hand, w_MeasuringCan be subject to significant errors, and in particular to systematic errors due to the presence of wind turbines or wind turbine blades.

For measuring w if it is determined that the wind turbine is operating under normal conditions, i.e. not in a significant stall condition and not in a disturbance condition_MeasuringCan be used from w_EstimatingThe obtained information is calibrated to take into account the influence w_MeasuringThe systematic error of (2).

In order to eliminate or at least mitigate the influence w_MeasuringIn the calibration phase, w_MeasuringValue of (a) and s_{Turbine wheel}Can be at different time points t_1,t₂,…,t_nThe measurement is repeated. In this case, let p be assumed_{Turbine wheel}At a time point t_1,t₂,…,t_nAre also known. Then, in some embodiments, the sequence of ordered pairs S is determined to be

Wherein, w_Measuring(t_i) Indicating at a point in time t_iPoint w_MeasuringValue of (a), s_{Turbine wheel}(t_i) Indicating at a point in time t_iAt s_{Turbine wheel}A value of and p_{Turbine wheel}(t_i) Indicating at a point in time t_iAt p_{Turbine wheel}I is 1, …, n.

Let p be_{Turbine wheel}It is known that this formula is abbreviated to

And for even greater simplicity, this formula is written as

Wherein w_Estimating(t_i)＝w_Estimating(s_{Turbine wheel}(t_i))＝w_Estimating(s_{Turbine wheel}(t_i),p_{Turbine wheel}(t_i))，i＝1,…,n。

In some embodiments, the time point t_iTime intervals of fixed or variable length may be identified, and w_Measuring(t_i) May be at t_iAn average measured wind speed over the identified time interval. For example, w_Measuring(t_i) Can be at t_iDuring the relevant interval (e.g. at interval t)_i-Δ_t,t_i]Period of time, wherein_tA predetermined time delay) of the wind speed. For example, w_Measuring(t_i) May be the instantaneous wind speed at the point in time t_iMoving averages of (1), e.g. simple moving averages or exponential moving averagesAnd (4) average number. It is intended that in these embodiments, w_Estimating(t_i) Can also be at t_iWithin an identified time interval (e.g., at interval t)_i-Δ_t,t_i]Inner) of the wind turbine, and/or w_Estimating(t_i) It may also be a moving average, such as for example a simple moving average or an exponential moving average, in particular having and by w_Measuring(t_i) The moving averages for the same or similar sampling windows are identified.

In some embodiments, a sequence S of ordered pairs may be used in order to determine a relationship between measured and estimated wind characteristics at a wind turbine location. The relationship may be, for example, a transfer function, and may be stored, for example, in a memory of a local controller or processor of the wind turbine, or elsewhere.

In some embodiments, the relationship between measured wind characteristics and estimated wind characteristics may be based on w_MeasuringMeasured value of and w_EstimatingIs calculated (based at least in part on the state s of the wind turbine)_{Turbine wheel}) Obtained by other means (e.g. at least partly using interpolation and/or regression analysis and/or monte carlo methods). In some embodiments, interpolation and/or regression analysis and/or monte carlo methods may be based on S.

In some alternative embodiments, the relationship between measured and estimated wind characteristics may be based in a similar manner and in particular on a sequence of ordered pairs obtained as described

To obtain, but wherein the wind characteristic w is measured_Measuring(t_i) And a measured state s of the wind turbine_{Turbine wheel}(t_i) And parameter p_{Turbine wheel}(t_i) Related to the same type of wind turbine among the wind turbines. Thus, in some embodiments, w_Measuring(t_i) And w_Estimating(t_i) Related to the same type of wind turbine among the considered wind turbines, and the relation S is based on the same type of wind turbine. Therefore, in the first of the present disclosureIn some embodiments, the S-based relationship is based on the same type of wind turbine.

In still further alternative embodiments, the sequence of ordered pairs S may be obtained by a simulation of the wind turbine, and thus the relationship between the measured and estimated wind characteristics based on S may be obtained by the simulation.

It is intended that the value in S is not based on a significant stall condition or disturbance condition of the wind turbine, i.e. for all time points or time intervals t_1,t₂,…,t_nNeither wind turbine is in a significant stall condition or disturbance condition. In embodiments in which S is obtained taking into account the same type of wind turbine among the wind turbines, it is intended that for all points in time or time intervals t_1,t₂,…,t_nNone of the wind turbines of the same type are in a significant stall condition or disturbance condition. In embodiments where S is obtained by simulation, no significant stall or disturbance conditions of the wind turbine are simulated, and for all simulated time points or time intervals t_1,t₂,…,t_nNeither a significant stall condition or a disturbance condition of the wind turbine is simulated.

With time t in the calibration phase under various wind characteristics_1,t₂,…,t_nAn increase in the number n of ordered pairs in the sequence S, and for each possible output ω of the sensor for measuring the wind characteristic at the wind turbine location, e.g. for each possible output ω of the anemometer, typically some pairs in the sequence S have ω as a first component, or have a first component close to ω. In some embodiments, interpolation or regression may alternatively be used to obtain missing data.

Order to

Is w_MeasuringA set of points in time equal or close to ω, where ω is the possible output of a sensor (e.g., a local anemometer) used to measure wind characteristics. Set T [ omega ]]Is an ordered set and can be written as T [ omega ]]＝{t_ω,1,t_ω,2,…}。

Symbol

Equal or approximately equal is indicated, wherein two scalars or vectors are considered equal or approximately equal if their distance according to a suitable metric is below a fixed limit. The fixed limit may be determined based on a characteristic of a sensor used to measure a characteristic of the wind, for example, based on a variance affecting an output of the sensor and/or based on a tolerance of a component or part included in the wind turbine.

Let S [ omega ] be a subsequence of S containing exactly those pairs in S whose first component is equal to omega or close to omega. For sufficiently large n, the subsequence S [ omega ] is desirably non-empty, and will contain, in an ordered manner, all ordered pairs in the sequence S having, as a first component, a value equal to or close to omega, i.e. all ordered pairs in the sequence S

S[ω]＝((w_Measuring(t_ω,1),w_Estimating(t_ω,1)),(w_Measuring(t_ω,2),w_Estimating(t_ω,2)),…)

Wherein

Then, the expected value E_EstimatingOmega and S omega]Is associated as a slave S [ omega ]]The obtained sequence S_Estimating[ω]：＝(w_Estimating(t_ω,1),w_Estimating(t_ω,2) …) by replacing S [ omega ] by its second component]Ordered pairs of (1). Expected value E_Estimating(ω) can be, for example, the sequence (w)_Estimating(t_ω,1),w_Estimating(t_ω,2) …) or a geometric mean or median. In some alternative embodiments, E_Estimating(ω) can be obtained from S by interpolation or regression.

Thus, the expected value E is when the output value of a wind characteristic sensor measuring the wind characteristic at the wind turbine location is equal to or close to ω, e.g. when the anemometer outputs ω or a value close to ω_Estimating(omega) form w_EstimatingIs calculated from the expected value of (c).Expected value E_Estimating(ω) is a desired wind characteristic value of the estimated wind characteristic determined from the measured wind characteristic and determined based on the sequence S. Symbol E_EstimatingAn expected value indicative of the estimated wind characteristic is indicated. Thus, E indicates desire. In an alternative embodiment using the same type of wind turbine, E_Estimating(ω) is related to the same type of wind turbine, i.e. ω refers to the possible output of e.g. an anemometer or a wind characteristics sensor of the same type of wind turbine. In an alternative embodiment, in which the wind turbine is simulated, E_Estimating(ω) is related to the simulated wind turbine, i.e., ω refers to the possible output of the simulated wind turbine, e.g., a simulated anemometer or a simulated wind characteristics sensor.

S and/or S [ omega ]]And/or S_Estimating[ω]And/or E_Estimating(ω) may be stored as a function of ω using any suitable data structure and on any suitable device and/or medium and/or by using any suitable system. In particular, vectors containing pairs or vectors containing scalar pairs or vector pairs or lists containing scalar pairs or vector pairs or hash tables or any nested combination of said data structures may be used, wherein said data structures may be stored remotely or locally at the wind turbine location on and manipulated by any suitable memory or computer or medium. The relevant data may be transmitted over, for example, a network or transmission line, one or more cables and/or one or more waveguides, or by using a wireless communication system. The data structure may be stored permanently or only for a required time interval, for example once E is obtained, for example in a calibration phase considering a wind turbine or a wind turbine of the same type or a simulated wind turbine_Estimating(ω) examples, implementation of S and/or S [ ω ] may be eliminated]And/or S_Estimating[ω]The example of (1) data structure.

It is intended that E_EstimatingA transfer function may be formed from wind speed measured using a wind characteristic sensor (e.g., an anemometer mounted on the wind turbine) to a desired wind speed estimated from turbine behavior. In an alternative embodiment using the same type of wind turbine, E for the same type of wind turbine is assumed_EstimatingEqual or close to E to be obtained on the actual physical wind turbine_EstimatingThe result of (1). In an alternative embodiment where a wind turbine is simulated, assume E for the simulated wind turbine_EstimatingEqual or close to E to be obtained on the actual physical wind turbine_EstimatingThe result of (1).

The transfer function E takes into account the fact that the wind turbine is known to operate at or near optimum (e.g., during wind turbine verification) and the fact that significant stall and disturbance conditions are not present_EstimatingCan be generated in particular from S.

In some embodiments of the present disclosure, in the calibration phase, one or more transfer functions E may be obtained_{Estimation, 1},E_{Estimation, 2},…,E_{Estimate v}Where v ≧ 1, the one or more transfer functions form a finite sequence of transfer functions

E_{Estimate, SEQ}＝(E_{Estimation, 1},E_{Estimation, 2},…,E_{Estimate v})。

Sequence E_{Estimate, SEQ}Some of the transfer functions in (a) may be based on measurements related to the wind turbine, E_{Estimate, SEQ}Some of the other transfer functions in (b) may be obtained based on measurements related to the same type of wind turbine among the wind turbines. E_{Estimate, SEQ}Further transfer functions in (b) may be obtained based on a simulation of the wind turbine, i.e. by replacing the measured values by a simulation based on e.g. a physical model of the wind turbine. And, when, for example, the calibration phase is repeated, the sequence E_{Estimate, SEQ}Different relationships in (a) may be associated with different time periods. Different transfer functions (for example transfer functions obtained using measurements and/or simulations, for example relating to wind turbines or wind turbines of the same type) may be combined so as to form a sequence E_{Estimate, SEQ}Of the transfer function. The combination may be based on averaging, weighted averaging, interpolation, etc., for example. In addition, the sequence E is obtained_{Estimate, SEQ}For any value ω in the domain of the transfer function in the sequence, for example by averaging

E_Estimating(ω)＝AVERAGE((E_{Estimation, 1}(ω),E_{Estimation, 2}(ω),…,E_{Estimate, v}(ω)))

From sequence E_{Estimate, SEQ}Obtaining an overall transfer function E_EstimatingAnd wherein AVERAGE may indicate any AVERAGE, such as a weighted AVERAGE, where, for example, a more recently obtained transfer function obtains more weight in the averaging operation. AVERAGE may also indicate, for example, an arithmetic or geometric mean or a median. In some embodiments, for some ω, missing data may be obtained, for example, by interpolation or regression.

In some embodiments, E_Estimating(ω) identifies the transfer function, i.e. the relation between the measured wind characteristics and the desired estimated wind characteristics. If and only if w_Estimating＝E_Estimating(w_Measuring) When this is true, the value w of the wind characteristic is estimated_EstimatingAnd consider E_EstimatingMeasured value w of wind characteristics of_MeasuringAnd (4) correlating. It is intended that the relationship E_EstimatingA transfer function is formed. It is therefore intended that the transfer function E_EstimatingIdentifying measured wind characteristics w_MeasuringAnd estimating wind characteristics w_EstimatingThe relationship between them.

In some embodiments, the marker measures a wind characteristic w_MeasuringAnd estimating wind characteristics w_EstimatingTransfer function E of the relationship between_EstimatingBased on w by other means (e.g. using interpolation and/or regression analysis and/or monte carlo methods, at least in part)_MeasuringMeasured value of and w_EstimatingBased at least in part on the state s of the wind turbine_{Turbine wheel}) To obtain the final product.

In some alternative embodiments, E_EstimatingAlternatively obtained under consideration of the same type of wind turbine or a simulated wind turbine.

Determining or adjusting 200 one or more wind characteristic relationships (i.e., E) when it is known that there is no significant stall condition or disturbance condition of the wind turbine_EstimatingAnd/or E_{Estimate, SEQ}) Is performed so that during calibration, w_EstimatingClose to the actual value of the wind characteristic at the wind turbine. When a relationship is determined or adjusted according to block 206, the relationshipCan be E_EstimatingOr sequence E_{Estimate, SEQ}The relationship (2).

In some embodiments, in the calibration phase, the wind turbine with the wind speed estimator as part of its controller software may be performed as desired at a known estimator (i.e., w_EstimatingClose to the actual wind conditions at the location of the wind turbine). For example, blade cleaning is ensured and, for example, an anemometer forming a wind characteristic sensor of a wind turbine works properly. During the calibration phase, data from, for example, a turbine anemometer (i.e., w)_Measuring) And data from the wind speed estimator (i.e. w)_Estimating) Is collected and w is calculated_MeasuringValue of (a) and w_EstimatingIs a transfer function E between the values of_EstimatingE.g. as described above for some embodiments of the disclosure. Transfer function E_EstimatingAllowing calculation of the desired output from the estimator, i.e. the desired E, based on the wind speed measured by e.g. an anemometer_Estimating(w_Measuring) Is close to w_Estimating。

FIG. 3 illustrates operational stages 300 of a method for operating a wind turbine according to some embodiments of the present disclosure, the operational stages including: measuring 302 a wind characteristic with a wind characteristic sensor (e.g., with wind characteristic sensor 58), thereby obtaining wind characteristic data; a state of the wind turbine is measured 304 with at least one wind turbine state sensor and an estimated wind characteristic is determined from the measured state of the wind turbine and a parameter of the wind turbine.

The operational phase 300 further includes comparing 306 the estimated wind characteristics to expected wind characteristics determined from the measured wind characteristics, where the expected wind characteristics are based on one or more wind characteristic relationships (i.e., E.E._EstimatingAnd/or E_{Estimate, SEQ}) Determined, for example, based on determining or adjusting 200 one or more relationships between wind characteristics in one or more calibration phases, as indicated, for example, by block 206. The operational phase 300 further includes operating or shutting down 308 the wind turbine based at least in part on the comparison. Determining or adjusting 206 relationships as described in FIG. 2 may form E_EstimatingOr sequence E_{Estimate, SEQ}Is determined therefrom as the measured wind characteristic w measured according to block 302_MeasuringThe desired wind characteristics of the function of.

During an operational phase, i.e. in particular when no calibration phase is performed, when a wind characteristic sensor output value w for measuring a wind characteristic is used_MeasuringWhen, e.g., based on the relation E_EstimatingOr sequences E obtained in one or more calibration phases_{Estimate, SEQ}Determined desired value E_Estimating(w_Measuring) Give w_EstimatingIs calculated from the expected value of (c).

Thus, w_MeasuringExpected value E of_Estimating(w_Measuring) Function mitigation direct impact w_MeasuringAnd at least if there are no significant stall conditions and disturbance conditions of the wind turbine, the desired value E_Estimating(w_Measuring) Approximating w as described close to the true wind characteristics at the wind turbine_EstimatingAnd w_MeasuringTypically directly affected by significant error and significantly different from w_EstimatingAnd thus is different from the true wind characteristics.

After calibration, during an operational phase, the wind characteristics w are measured_MeasuringFor obtaining wind characteristics w from measurements_MeasuringA determined value of the desired wind characteristic, the desired value being based on E_EstimatingE of (A)_Estimating(w_Measuring). Whenever, for example, E_Estimating(w_Measuring) Is significantly different from w_EstimatingWhen this happens, undesirable things may occur, and in particular a significant stall condition or disturbance condition of the wind turbine may occur, which makes these two values significantly different.

Thus, during the operating phase, based on e.g._Estimating(w_Measuring) And w_EstimatingIs beneficial to operate the wind turbine. In particular, the operating phase may be followed by one or more calibration phases. When for example E_Estimating(w_Measuring) Is close to w_EstimatingWhen a significant stall condition or disturbance condition of the wind turbine may not exist, and the wind turbine is in accordance with w_Estimating(i.e. according to w)_Estimating＝w_Estimating(s_{Turbine wheel})＝w_Estimating(s_{Turbine wheel},p_{Turbine wheel}) Operate because of the state s of the wind turbine_{Turbine wheel}And/or according to the state s of the wind turbine_{Turbine wheel}Parameter p with wind turbine_{Turbine wheel}Wind characteristic w estimated together with wind characteristic measured by wind characteristic sensor_MeasuringMore accurate and also more desirable wind characteristics (e.g. E)_Estimating(w_Measuring) Is more accurate.

In some embodiments, if wind characteristics are desired (e.g., E)_Estimating(w_Measuring) ) is significantly different from w_EstimatingE.g. when e.g._Estimating(w_Measuring) And w_EstimatingWhen the value of the magnitude of the difference between is above a predetermined threshold value (e.g. a threshold value between 0.5m/s and 2 m/s), then the wind turbine may be in a significant stall condition or in a disturbance condition, and thus the wind turbine may be in accordance with e.g._Estimating(w_Measuring) Operate because of the value w_EstimatingTypically unreliable and inaccurate in this case, based on w_MeasuringE of (A)_Estimating(w_Measuring) Possibly more precisely. Alternatively and/or depending on the magnitude of the difference, the wind turbine may be completely shut down in order to prevent possible damage to the wind turbine. If E is_Estimating(w_Measuring) Is significantly different from w_EstimatingThe wind turbine may thus be shut down or stopped or operated in a very conservative manner to prevent damage (e.g., when E_Estimating(w_Measuring) And w_EstimatingThe value of the magnitude of the difference between is higher than a predetermined threshold (e.g., a threshold between 0.5m/s and 2 m/s). In some embodiments, the threshold may be any value greater than, for example, 0.5 m/s.

In some embodiments, during operation of the wind turbine after the calibration phase has been completed, the turbine constantly calculates the desired wind speed E at regular intervals_Estimating(w_Measuring) (e.g., in real time or near real time, or, e.g., once per hour, once per day, or once per week). According to an embodiment of the present disclosure, if an estimated wind speed E is desired_Estimating(w_Measuring) And a model-based estimated wind speed w_EstimatingPhase differenceBeyond a certain threshold (e.g., a threshold between 0.5m/s and 2 m/s), several actions may be taken. In one embodiment, the turbine controller of the wind turbine switches to use e.g. from the local anemometer according to w_MeasuringThe desired estimated wind speed E obtained_Estimating(w_Measuring) Rather than a model-based estimate w_EstimatingAs an input to the main controller. In some embodiments, a message will be generated indicating that the turbine needs to be checked. In some embodiments, the turbine will switch to a safer mode of operation that protects the turbine from potential damage due to certain conditions (such as increased pitch angle) to avoid stall. In some embodiments, the expected value E will be_Estimating(w_Measuring) And the actual value w_EstimatingThe non-matching patterns therebetween are compared to pre-calculated or pre-determined failure patterns stored in software or memory associated with the wind turbine or wind turbine controller, and in some embodiments, action is taken based on the particular failure pattern.

More generally, embodiments of the present disclosure relate to a method for operating a wind turbine comprising a sensor for measuring a wind property w_MeasuringAnd for measuring a state s of the wind turbine_{Turbine wheel}The method comprises: determining or adjusting one or more wind characteristic relationships, i.e. relationship E_EstimatingOr the sequence of the relation E_{Estimate, SEQ}(ii) a And, performing an operation phase, the operation phase comprising: measuring the wind characteristics with a wind characteristics sensor to obtain a measured wind characteristic w_Measuring(ii) a Measuring the state s of a wind turbine using at least one wind turbine state sensor_{Turbine wheel}And determining an estimated wind characteristic w based on the measured state of the wind turbine and the parameters of the wind turbine_Estimating(w_Estimating＝w_Estimating(s_{Turbine wheel})＝w_Estimating(s_{Turbine wheel},p_{Turbine wheel}) B) of the group A and B); will estimate the wind characteristics w_EstimatingAnd based on the measured wind characteristic w_MeasuringDetermined desired wind characteristics E_Estimating(w_Measuring) A comparison is made in which the wind characteristics E are expected_Estimating(w_Measuring) Based on one or more wind characteristic relationships (i.e. based on E)_EstimatingOr based on E_{Estimate, SEQ}) To determine; and, operating or shutting down the wind turbine based at least in part on the comparison.

For example, if the wind characteristics E are desired_Estimating(w_Measuring) Sequence E based on relationships obtained, for example, taking into account wind turbines and/or wind turbines of the same type and/or simulations_{Estimate, SEQ}＝(E_{Estimation, 1},E_{Estimation, 2},…,E_{Estimate, v}) Then the wind characteristics E are expected_Estimating(w_Measuring) Can be obtained by averaging:

E_estimating(w_Measuring)

＝AVERAGE((E_{Estimation, 1}(w_Measuring),E_{Estimation, 2}(w_Measuring),…,E_{Estimate, v}(w_Measuring)))

In some embodiments, interpolation or regression may be used.

In some embodiments, one or more wind characteristic relationships are determined or adjusted (i.e., E is determined or adjusted)_EstimatingOr sequence E_{Estimate, SEQ}One or more relationships E in_{Estimate i}And thus determine or adjust the sequence E_{Estimate, SEQ}) Is performed when the wind turbine is not in a significant stall condition and is not in a disturbance condition, and comprises: measuring wind characteristics w of a wind turbine using a wind characteristics sensor of the wind turbine_MeasuringThereby obtaining a measured wind characteristic of the wind turbine; measuring the state s of a wind turbine using at least one wind turbine state sensor_{Turbine wheel}And determining an estimated wind characteristic (w) of the wind turbine from the measured state of the wind turbine and the parameters of the wind turbine_Estimating＝w_Estimating(s_{Turbine wheel},p_{Turbine wheel}) B), determining or adjusting a relation E between a measured wind characteristic of the wind turbine and an estimated wind characteristic of the wind turbine_EstimatingOr E_{Estimate i}Wherein i indicates the ith relationship that is currently determined or adjusted; and adjusting one or more wind characteristic relationships, i.e. E_EstimatingOr E_{Estimate, SEQ}To include measured wind characteristics of the wind turbine and estimated wind characteristics of the wind turbineRelationship between sexes E_EstimatingOr E_{Estimate i}。

In some embodiments, one or more wind characteristic relationships (i.e., E) are determined or adjusted_EstimatingOr E_{Estimate, SEQ}) The method comprises the following steps: operating the same type of wind turbine as the wind turbine when the same type of wind turbine is not in a significant stall condition and is not in a disturbance condition, the same type of wind turbine including a wind characteristic sensor and at least one wind turbine condition sensor; during operation of the same type of wind turbine, measuring a wind characteristic of the same type of wind turbine with a wind characteristic sensor of the same type of wind turbine, thereby obtaining a measured wind characteristic of the same type of wind turbine; and measuring a state of the same type of wind turbine with at least one wind turbine state sensor of the same type of wind turbine and determining an estimated wind characteristic of the same type of wind turbine from the measured state of the same type of wind turbine and a parameter of the same type of wind turbine; determining or adjusting a relationship between measured wind characteristics of wind turbines of the same type and estimated wind characteristics of wind turbines of the same type; and adjusting one or more wind characteristic relationships (i.e., E)_EstimatingOr E_{Estimate, SEQ}) To include a relationship between measured wind characteristics of the same type of wind turbine and estimated wind characteristics of the same type of wind turbine.

In some embodiments, one or more wind characteristic relationships (i.e., E) are determined or adjusted_EstimatingOr E_{Estimate, SEQ}) The method comprises the following steps: simulating wind and wind turbine operation for the wind turbine without significant stall and disturbance conditions of the wind turbine, the simulation based at least in part on a model of the wind turbine; obtaining simulated wind characteristics, a simulation state and simulation parameters of the wind turbine, and determining simulated estimated wind characteristics according to the simulation state of the wind turbine and the simulation parameters of the wind turbine; determining or adjusting a relationship between the simulated wind characteristic and the simulated estimated wind characteristic; and adjusting one or more wind characteristic relationships (i.e., E)_EstimatingOr E_{Estimate, SEQ}) To include a relationship between simulated wind characteristics and simulated estimated wind characteristics.

In some embodiments, one or more wind characteristic relationships (i.e., E)_EstimatingOr E_{Estimate, SEQ}) Further combined into a single combined relationship and the desired wind characteristics are based on the single combined relationship. For example, for the relation E obtained for example under consideration of wind turbines and/or wind turbines of the same type and/or simulations_{Estimate, SEQ}＝(E_{Estimation, 1},E_{Estimation, 2},…,E_{Estimate, v}) Sequence E of_{Estimate, SEQ}A single combinatorial relationship may associate an average value with each value ω in the domain of relationships in the sequence (E)_Estimating(ω)＝AVERAGE((E_{Estimation, 1}(ω),E_{Estimation, 2}(ω),…,E_{Estimate, v}(ω)))). In some embodiments, missing data for some ω may be obtained, for example, using interpolation or regression.

In some embodiments, data originating from wind characteristic sensors (e.g. anemometers) and estimated wind speeds are continuously evaluated and compared during normal turbine operation, i.e. during an operational phase, using, for example, statistically determined transfer functions. If desired, estimating the wind characteristics E_Estimating(w_Measuring) And estimating wind characteristics w_EstimatingIf a match between them is not available with predetermined requirements, then it is assumed that the wind turbine is not operating as expected (e.g. due to icing, blade fouling or stalling) and a message is generated to the remote control centre so that appropriate steps can be taken to remedy the problem.

FIG. 4 summarizes details related to methods for operating a wind turbine according to some embodiments of the present disclosure. FIG. 4 shows that wind characteristic sensor 402, such as wind characteristic sensor 58, may be provided with a measured wind characteristic w, as indicated by 406_MeasuringAnd at least one sensor 404 measuring a state of the wind turbine provides a state s of the wind turbine as indicated by 408_{Turbine wheel}Wherein the condition may, for example, comprise a rotational speed or torque of a rotor and/or a shaft (e.g., rotor shaft 44) of the wind turbine and/or comprise a power output of, for example, a generator. It is assumed that parameters 410 of the wind turbine are known, for example, that the pitch angle of the blades of the wind turbine and/or the configuration of the gearbox are known. Parameter 410 utilizationP as indicated by 412_{Turbine wheel}And (4) indicating. By using the physics model 416, the wind characteristics w are estimated_EstimatingObtained (as shown by 420) as a function of the state 408 of the wind turbine and the parameters 412 of the wind turbine through a physics model 416. From the measured wind characteristics w indicated by 406_MeasuringBased on one or more relationships (i.e., E) as indicated schematically by 414_EstimatingOr E_{Estimate, SEQ}) To obtain desired values of estimated wind characteristics, e.g. E_Estimating(w_Measuring). Expected value E_Estimating(w_Measuring) Indicated by 418. At desired estimated wind characteristics E indicated by 418_Estimating(w_Measuring) And an estimated wind characteristic w indicated by 420_EstimatingA comparison 422 is performed. As indicated by 424, based at least in part on the comparison 422, the wind turbine ultimately operates or performs a shutdown. Wind turbine operation or shutdown 424 may be based on E_Estimating(w_Measuring) And w_EstimatingA comparison 422 therebetween, and in particular wind turbine operation may further depend on E_Estimating(w_Measuring) And/or w_EstimatingAnd/or dependent on e.g. E_Estimating(w_Measuring)、w_EstimatingBased on the results of the comparison 422.

The method of the present disclosure involves calibrating a wind characteristics sensor, i.e., a wind measuring device, w during normal operating time, taking into account a physical model_EstimatingObtained from the physical model. Wind characteristic sensors (i.e. wind measuring devices at the location of the wind turbine) are used to detect insufficient performance of the wind turbine and/or inappropriate behaviour of the wind turbine (e.g. due to a significant stall condition or disturbance condition of the wind turbine, in particular when estimating the value w_EstimatingBecomes inaccurate, i.e., when the wind speed estimation algorithm based on the physical model no longer works properly).

The method of the present disclosure is particularly beneficial for ice detection, stall detection, possibility to perform seasonal calibration phase. Due to E_Estimating(w_Measuring) In addition, it is possible to obtain a reliable power curve with a nacelle anemometer forming, for example, a wind characteristic sensor of a wind turbineAnd (6) measuring.

In some embodiments of the present disclosure, a method for operating a wind turbine is described, wherein e.g. is used_Estimating(w_Measuring) And w_EstimatingThe comparison therebetween shows the estimated wind characteristics w_EstimatingIs significantly different from measuring wind characteristics w_MeasuringDetermined value of desired wind characteristic (e.g. E)_Estimating(w_Measuring) In accordance with measured wind characteristics w_MeasuringDetermined desired wind characteristics E_Estimating(w_Measuring) To operate or shut down.

In some embodiments, comparing 306, 422 may include comparing based on one or more relationships (i.e., E)_EstimatingOr E_{Estimate, SEQ}) To obtain an estimated wind characteristic w_EstimatingAnd expected wind characteristics E_Estimating(w_Measuring) The difference between them delta. It is intended that Δ may be Δ ═ w_Estimating-E_Estimating(w_Measuring) In which E_EstimatingMay be the relation E_EstimatingOr based on the sequence E_{Estimate, SEQ}. In some embodiments, based on sequence E_{Estimate, SEQ}(wherein E_{Estimate, SEQ}＝(E_{Estimation, 1},E_{Estimation, 2},…,E_{Estimate, v}) A) difference Δ may be Δ ═ w_Estimating-AVERAGE((E_{Estimation, 1}(w_Measuring),E_{Estimation, 2}(w_Measuring),…,E_{Estimate, v}(w_Measuring) AVERAGE) where AVERAGE may indicate any convenient AVERAGE. Sequence E_{Estimate, SEQ}Relation E in_{Estimate i}May be obtained under consideration of a wind turbine or a wind turbine of the same type or by simulation. For consistency of the symbols, it is still written as: e_Estimating(w_Measuring)＝AVERAGE((E_{Estimation, 1}(w_Measuring),E_{Estimation, 2}(w_Measuring),…,E_{Estimate, v}(w_Measuring))). In some embodiments, the wind turbine operates based at least in part on the magnitude of the difference Δ.

In some embodiments, comparison 422 may correspond to comparison 306, and includes obtaining estimated wind characteristics w_EstimatingAnd expected wind characteristics (e.g. E)_Estimating(w_Measuring) A difference Δ between, e.g. a difference w_Estimating-E_Estimating(w_Measuring) And operating the wind turbine based at least in part on the magnitude of the difference delta.

The difference Δ may be a scalar or vector, and the magnitude of the difference (e.g., difference w)_Estimating-E_Estimating(w_Measuring) By any suitable metric or norm, particularly by, for example, a euclidean norm, maximum norm, etc. In particular, the difference Δ (e.g., w)_Estimating-E_Estimating(w_Measuring) Is intended to be non-negative and real, and if and only if a scalar or vector operand (e.g., w)_EstimatingAnd E_Estimating(w_Measuring) When equal, the magnitude of the difference Δ is zero.

In some embodiments of the present disclosure, the magnitude of the difference Δ (e.g., the difference w) is determined as the magnitude of the difference Δ_Estimating-E_Estimating(w_Measuring) Is below a first threshold value (e.g. below 2m/s or below 1m/s), the wind turbine is based on the estimated wind characteristic w_EstimatingTo operate.

For example, in some embodiments, when w_EstimatingIs close to E_Estimating(w_Measuring) Time, difference w_Estimating-E_Estimating(w_Measuring) Becomes close to zero and therefore when w_EstimatingIs close to E_Estimating(w_Measuring) Time difference value w_Estimating-E_Estimating(w_Measuring) Is below a first threshold. Under such conditions, a significant stall condition or disturbance condition of the wind turbine is undesirable, and therefore the wind turbine is dependent on w_EstimatingOperation (especially when w_EstimatingRatio of possible E_Estimating(w_Measuring) And/or w_MeasuringMore precisely).

In some embodiments, the magnitude of the difference Δ (e.g., the difference w) is determined as the magnitude of the difference Δ_Estimating-E_Estimating(w_Measuring) Magnitude) is above a first threshold, the wind turbine is based on the desired wind characteristic E_Estimating(w_Measuring) To operate. For example, when w_EstimatingIs significantly different from E_Estimating(w_Measuring) When w_Estimating-E_Estimating(w_Measuring) Is increased above the first threshold value, and the wind turbine may be in a stall or disturbance condition, and thus w_EstimatingMay become unreliable and inaccurate. Thus, according to the desired wind characteristic value (e.g. according to E)_Estimating(w_Measuring) Operating the wind turbine is beneficial for wind turbine operation and/or for safe operation of the wind turbine in order to prevent damage and/or for shutdown of the wind turbine.

In some embodiments, when the difference Δ (e.g., the difference w)_Estimating-E_Estimating(w_Measuring) ) is above a second threshold (e.g. above 3m/s), the turbine switches to a safe operation mode.

The safe mode may be related to the control of one or more pitch angles of one or more blades of the wind turbine in order to prevent a significant stall condition, or the safe mode may include completely shutting down the wind turbine.

In some embodiments, when the difference Δ (e.g., the difference w)_Estimating-E_Estimating(w_Measuring) ) is above the first and/or second threshold, a message is transmitted to the operator.

The transmission may be fully automated, and the operator may be one or more human operators and/or one or more computers or fully or partially automated systems configured to control the wind turbine. An operator or one or more computers or fully or partially automated systems configured to control the wind turbine may be located in the wind farm or at a remote location, or even at the wind turbine location or in the wind turbine itself. The message may be transmitted by any suitable means, such as digital packets over a network, or as a modulated radio wave signal, or over a cable or optical waveguide. The message may include any additional information useful for control of the wind turbine and/or obtaining information about the wind turbine state or condition.

In some embodiments, the difference Δ (e.g., difference w)_Estimating-E_Estimating(w_Measuring) The size of the sequence is memorized at different points in time forming the sequence, and normality is determined based on the sequenceA condition or a significant stall or disturbance condition, wherein in case of a significant stall or disturbance condition, the type of fault is determined according to the sequence and the wind turbine is operated according to the determined type of fault.

Memorizing the difference value delta (e.g. difference value w) at different time points_Estimating-E_Estimating(w_Measuring) For example, periodically sampling and storing the difference values, produces a sequence of values that form a history of difference values. Based on the history, it is for example possible to record the difference Δ (e.g. the difference w) when e.g. a significant stall condition or a disturbance condition occurs_Estimating-E_Estimating(w_Measuring) How the size increases. From the history of the difference Δ, information about the type of fault may be obtained, wherein the type of fault may for example specify whether a significant stall condition is occurring or what fault is occurring in different possible situations, e.g. stating whether the blade is icing or whether dust or ageing may be affecting the operation of the wind turbine.

Other sources of information may also be used to determine the type of fault, such as information obtained from thermometers and/or other sensors placed at or around the location of the wind turbine, for example. The information sources may also include weather forecasts or observations and wind forecasts or measurements at different locations, including wind turbine locations.

In some embodiments, based on the estimated wind characteristics w_EstimatingAnd based on the measured wind characteristic w_MeasuringDetermined value of desired wind characteristic (e.g. E)_Estimating(w_Measuring) In the event of a stall, the wind turbine is shut down or operated so as to control the pitch angle to avoid stalling of the wind turbine.

In some embodiments, operating or shutting down the wind turbine includes adjusting the pitch angle to avoid a significant stall condition of the wind turbine.

The calibration phase may include w_MeasuringAnd s_{Turbine wheel}Such that a sufficient number of ordered pairs are obtained to obtain e.g. E for each possible output value ω of the wind characteristic sensor_EstimatingA sufficiently accurate and precise value of (ω). When e.g. with the process from S [ omega ] as described previously]The obtained sequence S_Estimating[ω]Can be determined, there can be a sufficiently accurate and precise value (e.g., when considering the sequence S)_Estimating[ω]In the case of a sample from a monte carlo experiment for which a desired confidence interval width is required to achieve a desired confidence level).

When it is known that no significant stall or disturbance condition exists, a calibration phase for determining or adjusting one or more wind characteristic relationships may be performed. The determination may be fully automatic, e.g. automatically checking the temperature and wind at the wind turbine location and other conditions, such as the presence of dust, or the determination may be partly automatic or the result of manual monitoring. The calibration phase may include the use of a measurement instrument, which may be removed after the calibration phase is complete. The manual monitoring may be present during the calibration phase and not thereafter, or the calibration may be fully automatic.

The calibration phase and the operating phase may be alternated, e.g. periodically, in order to recalibrate, i.e. adjust e.g. E_EstimatingOr E_{Estimate, SEQ}In order to take account of, for example, ageing of the wind turbine or other time-varying properties of the wind turbine and/or to take account of modifications of the aerodynamic properties of the wind turbine location.

In some embodiments, the calibration phase is repeated until the desired wind characteristic value approaches the estimated wind characteristic with sufficient accuracy and precision.

In some embodiments, during the calibration phase, measuring the wind characteristics further comprises measuring data from one or more wind measuring bars positioned at a distance from the wind turbine.

Thus, in some embodiments, w_MeasuringMay be a vector comprising values obtained from at least one local anemometer and/or at least one measuring pole for measuring wind conditions at a distance from the wind turbine.

In some embodiments, a wind turbine is described, the wind turbine comprising: at least one wind measurement sensor; a wind turbine state sensor to measure a state of the wind turbine for estimating wind characteristics at a wind turbine location; a control system configured to control the wind turbine based at least in part on an input formed by the measured wind characteristic measured by the wind measurement sensor and the measured wind turbine state measured by the wind turbine state sensor, wherein the control system is configured to operate the wind turbine according to the method described in the present disclosure. It is assumed that the wind turbine parameters are known to the control system.

In some embodiments, the wind characteristic may be a wind speed or a magnitude of the wind speed, and the wind characteristic sensor measures the magnitude of the wind speed or the wind speed. In some embodiments, the wind characteristic sensor may measure the magnitude and direction of the wind speed. In some embodiments, the wind characteristic sensor may measure a vector describing wind speed. In some embodiments, the magnitude of wind speed may be measured in m/s.

In some embodiments, the wind turbine further comprises an information processing system and at least one communication channel configured to transmit information regarding the comparison of the estimated wind characteristics to the desired wind characteristics during the operational phase.

The transmitted information, for example, transmitted over the communication channel, may be used to control or monitor the operation of the wind turbine.

Operating a wind turbine according to the method described herein or as illustrated in FIG. 4 is particularly beneficial for detecting significant stall or disturbance conditions, and allowing operation of the wind turbine to minimize the risk of damage to the wind turbine and/or maximize the performance of the wind turbine (particularly during significant stall or disturbance conditions), such as a difference Δ (e.g., difference w)_Estimating-E_Estimating(w_Measuring) An increase in size) may be associated with, for example, dust or ice deposits on the blades of the wind turbine, or with any other disturbance conditions and/or significant stall conditions. The method of the present disclosure may allow for a difference Δ (e.g., difference w) based on, for example, memory at different time points in the sequence_Estimating-E_Estimating(w_Measuring) ) to detect a disturbance condition and a significant stall condition. The difference delta (e.g. the difference w)_Estimating-E_Estimating(w_Measuring) Increase in size of the product)The addition may also be associated with, for example, a significant stall condition, which may also be based on, for example, a difference Δ (e.g., difference w) learned at different points in time_Estimating-E_Estimating(w_Measuring) ) of the size of the sample. In such a case, i.e. if a significant stall condition is detected, it is beneficial to adjust the pitch angle of one or more blades of the wind turbine or to shut down the wind turbine in order to prevent e.g. damage to the wind turbine.

Claims (10)

1. Method for operating a wind turbine comprising a wind characteristic sensor for measuring a wind characteristic and at least one wind turbine status sensor for measuring a status of the wind turbine, the method comprising:

determining or adjusting one or more wind characteristic relationships; and

performing an operational phase, the operational phase comprising:

measuring the wind characteristic with the wind characteristic sensor, thereby obtaining a measured wind characteristic;

measuring the state of the wind turbine with the at least one wind turbine state sensor and determining an estimated wind characteristic from the measured state of the wind turbine and a parameter of the wind turbine;

comparing the estimated wind characteristic to an expected wind characteristic determined from the measured wind characteristic, wherein the expected wind characteristic is determined based on the one or more wind characteristic relationships; and

operating or shutting down the wind turbine based at least in part on the comparison.

2. The method of claim 1, wherein determining or adjusting one or more wind characteristic relationships is performed when the wind turbine is not in a significant stall condition and not in a disturbance condition, and comprises:

measuring the wind characteristic of the wind turbine with the wind characteristic sensor of the wind turbine, thereby obtaining a measured wind characteristic of the wind turbine;

measuring the state of the wind turbine with the at least one wind turbine state sensor and determining an estimated wind characteristic of the wind turbine from the measured state of the wind turbine and a parameter of the wind turbine,

determining or adjusting a relationship between the measured wind characteristic of the wind turbine and the estimated wind characteristic of the wind turbine; and

adjusting the one or more wind characteristic relationships to include the relationship between the measured wind characteristic of the wind turbine and the estimated wind characteristic of the wind turbine.

3. The method of claim 1 or 2, wherein determining or adjusting one or more wind characteristic relationships comprises:

operating a same type of wind turbine as the wind turbine when the same type of wind turbine is not in a significant stall condition and is not in a disturbance condition, the same type of wind turbine including a wind characteristic sensor and at least one wind turbine condition sensor; and, during the operation of the wind turbines of the same type, the method further comprises:

measuring a wind characteristic of the same type of wind turbine with the wind characteristic sensor of the same type of wind turbine, thereby obtaining a measured wind characteristic of the same type of wind turbine; and

measuring the state of the same type of wind turbine with the at least one wind turbine state sensor of the same type of wind turbine and determining an estimated wind characteristic of the same type of wind turbine from the measured state of the same type of wind turbine and a parameter of the same type of wind turbine;

determining or adjusting a relationship between the measured wind characteristics of the wind turbines of the same type and the estimated wind characteristics of the wind turbines of the same type; and

adjusting the one or more wind characteristic relationships to include the relationship between the measured wind characteristics of the same type of wind turbine and the estimated wind characteristics of the same type of wind turbine.

4. The method of any one of the preceding claims, wherein determining or adjusting one or more wind characteristic relationships comprises:

simulating wind and wind turbine operation for the wind turbine without significant stall and disturbance conditions of the wind turbine, the simulation based at least in part on a model of the wind turbine;

obtaining simulated wind characteristics, simulated states and simulated parameters of the wind turbine, determining simulated estimated wind characteristics from the simulated states of the wind turbine and the simulated parameters of the wind turbine;

determining or adjusting a relationship between the simulated wind characteristic and the simulated estimated wind characteristic; and

adjusting the one or more wind characteristic relationships to include the relationship between the simulated wind characteristic and the simulated estimated wind characteristic.

5. The method of any of claims 1 to 4, wherein the one or more wind characteristic relationships are further combined into a single combined relationship, and wherein the desired wind characteristic is based on the single combined relationship.

6. A method according to any of the preceding claims, wherein the wind turbine is operated or shut down according to the desired wind characteristic determined from the measured wind characteristic, when the comparison shows that the estimated wind characteristic differs significantly from the desired wind characteristic determined from the measured wind characteristic.

7. The method of any preceding claim, wherein the comparing comprises obtaining a difference between the estimated wind characteristic and the desired wind characteristic, and operating the wind turbine is based at least in part on a magnitude of the difference.

8. The method of claim 7, wherein the wind turbine operates based on the estimated wind characteristic when the magnitude of the difference is below a first threshold.

9. The method of claim 7 or 8, wherein the wind turbine operates based on the desired wind characteristic when the magnitude of the difference is above the first threshold.

10. Method according to any of claims 7 to 9, characterized in that when the magnitude of the difference is above a second threshold, the turbine switches to a safe operation mode or shuts down.

Images